Time for a shift. Behavioral approaches to online learning evaluation.

Digital learning materials are on the rise in higher education settings (Treve, 2021; Zawacki-Richter, 2021). This implies a growing need for research on how they should be designed to assist learning effectively (Sailer & Schlag, 2012). One factor under consideration are emotions, since they are proven to be influential factors on learning (Pekrun et al. 2017; Tnyg et al. 2017). Our project explores whether empirical behavioral methods can provide amore valid insight into realistic learning processes and aims to identify facilitating factors for learning with digital learning materials. Anger and sadness are the most frequent emotional facial expressions during the learning session, according to our analysis. Anger, in particular, can also be interpreted as an expression of concentration or effort. Eyemovement data show whether students recognize relevant information within the learning material. Differences in facial expressions for different types of information presentation can not be confirmed (H1), which could be due to methodological artifacts (i.e., analysis window for facial reactions, material). We found correlations between disgust and negative affect and surprise and positive affect for some conditions (H2). Future analyses should use the focus on relevant areas of interest as cues for simultaneous analyses of facial expressions.